To: US ATLAS Computing Advisory Panel

Chip Watson, P. Elmer, I. Fisk, R. Mount, T. Hemmick

From: S. Aronson

Re: Charge to the Panel for 5-6 Jan., 2006 Meeting at BNL.

21 December, 2005

We will be soliciting your comments and recommendations on a number of items that have changed since our last review.

The upcoming DOE/NSF Review of U.S. LHC Computing (Feb., 2006) will concentrate on plans, schedules, risks and contingency plans for development, deployment and operation of the US ATLAS computing infrastructure. This will cover the period FY06-FY10, with emphasis on FY07 and FY08. We will be asked to consider 2 funding scenarios:

Baseline: constrained by agency provided guidance

Reduced: 10% less than current guidance.

1. Project Organization/Management

Are the deliverables, priorities and scope of the project clear?

Are the management tools being used effectively? Is the management process effective? Is change control working?

The Research Operations phase of US ATLAS has begun and computing funding must be shared with maintenance and operations (M&O) for the construction project and upgrade R&D. Priorities are assigned to tasks (that include M&O and computing items). Are the priorities/strategy correct for this allocation? Is the proposed computing funding plan, adequate?

Does the project have the right connection to and communication with external grid projects: the Worldwide LHC Computing Grid (WLCG) and the Open Science Grid(OSG)?

Are the risks of shortfalls in other international software efforts understood?

Does the management have an adequate reserve mechanism to react to budget uncertainties?

Is there sufficient communication between U.S. CMS and U.S. ATLAS?

Does the management have adequate S&C resources and planning to accommodate new collaborators (both new institutes and new researchers)? Is there a model for the incremental cost?

Is there adequate infrastructure support from BNL as host lab?

2. Software and Analysis Support

Is the project paying attention to overall performance and architecture issues? Is the manpower well spent in this area?

How is progress compared to milestones shown last March 2005?

Have the long term priorities been established and is the resource allocation consistent with the priorities?

Are the interactions with LCG (POOL, SEAL) and ATLAS proper/working/at the right level?

Is the distributed software effort, including DDM, at the right level?

Is the PanDA software development on track?

Is the core software effort balanced so as to provide U.S. researchers effective participation in the physics of ATLAS?

Is the U.S. ATLAS Physics analysis model plan on course? Is there an adequate implementation plan?

3. Grids and T1/T2 facilities.

Are the interactions with other grid projects, WLCG/OSG coordinated properly?

Are we effectively managing/exploiting all the resources from outside the US ATLAS computing program?

Assess the performance of the U.S. on the latest data and service challenges. Does this performance scale to the full LHC needs? Is there feedback on any U.S. specific problems?

Are we using program funds effectively to integrate the grid middleware?

Is the proposed facility ramp-up plan adequate, given the ATLAS computing model and the U.S. researcher’s needs?

Have infrastructure and operating costs of the T1 and T2 centers been fully considered in the plans?

In particular, at T1: Space, power and HVAC costs?

4. Next generation funding initiatives

A large proposal is being prepared for the Open Science Grid.

Is the U.S. ATLAS effort involved in this at the right level?

How many FTE’s are needed from the operations program to allow ATLAS to effectively use OSG?

Should U.S. ATLAS totally rely on OSG deliverables?

While we use this review as a “dress rehearsal” for our upcoming DOE/NSF review, we want the CAP review to be much more than that. We would like use the CAP to obtain advice on these important issues. We hope you will report briefly to us at the end of the review on 6 Jan. then provide us with a written report shortly thereafter.

The final report from the last PCAP review can be seen here:

HYPERLINK "http://www.usatlas.bnl.gov/computing/meet/0502PcapReview/" http://www.usatlas.bnl.gov/computing/meet/0502PcapReview/

The final report from our last DOE/NSF full review:

HYPERLINK "http://www.usatlas.bnl.gov/computing/meet/0503AgencyReview/" http://www.usatlas.bnl.gov/computing/meet/0503AgencyReview/